Professional Documents
Culture Documents
A Guide To Thermal Analysis PDF
A Guide To Thermal Analysis PDF
Thermal Analysis
This guide starts from applications of thermal analysis and its role in
simulation driven design. Fundamental concepts and principles will
be introduced such as conduction, convection, radiation, linear and
nonlinear heat transfer, steady state and transient analysis, etc.
Finally we will discuss how to choose appropriate finite element
analysis software for thermal analysis; and introduce strength of
midas NFX for solving thermal problems.
Thermal Analysis for Analysis Driven Design
Efficiency and innovation are contradictory challenges in the sector of product development, and
both them must be satisfied in modern companies that strive to remain competitive. Simulation
driven design process was introduced and has largely replaced the traditional prototyping-testing
process in the last decade.
Companies can benefit from the cost and time efficiency of design verification tools based on finite
element methods, and concentrate main efforts on product innovation.
Prototyping
Prototyping Testing 1 Prototype
Many Prototypes Many Tests
Test
1 Test
Production Production
Cost: $ $ $ $ Cost: $
Figure 1
By introducing computer aided design verification tools in the early stage of the product design.
Time and cost wasted in the prototype - test - redesign cycle are significantly reduced.
Page 3
Thermal Analysis Application
Thermal related problems are challenges commonly faced by product design engineers. Some of
the problems include overheating, excessive thermal stresses, thermal effect on dimensional
stability, etc. Following are some applications in which thermal factors need to be considered.
NFX Application
Design of cooling fins for high temperature / high tension pump
Bearing area
From 74.55 to 65.53 Position
Temperature Temperature
(13.7% decreases) <Original> <Improved >
Figure 3
Page 3
In the application above, cooling fins were designed to prevent overheating of a pump which
operates under high temperature and high pressure environment.
Temperature ()
Input_Disc_Temp.
brakes, hydraulic pumps are to be evaluated 600
NFX_Output
500
during the design process.
400
300
In the right application, thermal analysis was 200
performed to choose the appropriate type of 100
brake fluid. The maximum temperature 0
0 300 600 900 1200 1500
transferred to brake fluid was predicted using
Time (s)
heat transfer analysis. And the boiling point of
Brake fluid temperature as per time
the chosen brake fluid needs to be higher to
avoid the vapor lock phenomenon.
Figure 4
Displacement Temperature
Temperature Gradient
Stress Heat Flux
Figure 5 Figure 6
Page 3
As illustrated in figure 5 - 6, thermal analysis are usually performed to find the temperature
distribution, temperature gradient and heat flux of the target model. The simulation approach is
particularly beneficial for solving thermal problems because firstly you can use the same CAD
model to perform thermal analysis as well as structural analysis. Secondly measuring
temperatures and temperature changes can be very difficult in a real test, especially when
temperatures inside small parts and assemblies are to be decided.
Figure 7
Conduction
Conduction describes heat flowing inside a body. Heat energy is transferred through the chained
vibration of molecules or neutrons from high temperature region to low temperature region.
THOT TCOLD
K: Thermal conductivity
A: Area of the wall
L: Wall Thickness
L
Figure 8
Regardless of the state of a material such as
solid, liquid or gas, conduction always occurs if a Heat conduction of a simple plate
temperature difference (temperature gradient)
100
exists within the object.
Moreover, thermal energy moves from the high-
temperature region to a low-temperature region.
The simple example in figure 9 demonstrates
typical analysis result of conduction in heat 20
transfer.
Figure 9
Convection
Convection describes the mechanism of heat exchange between an external face of a solid body
and the surrounding fluid such as air, steam, water, oil, etc. The convection coefficient strongly
depends on the medium (e.g. steam, water, oil).
TSurface
Figure 10
Convection can be further classified into 2 types: natural convection and forced convection.
Natural convection happens because of gravity difference between cold and hot fluids. While
forced convection can created by external force (cooling fan, wind, etc.).
Figure 11
Figure 12 is an example of forced convection. NFX Application
Cooling fan is installed in the PCB system and PCB Cooling Forced Convection
initiates the convection. In this application
temperature distribution, fluid flow velocity and
fluid flow path were determined. It practically
uses fluid analyses (CFD) coupled with transient
heat transfer to solve such forced convection
problem.
Figure 12
Radiation
Radiation is the heat transfer phenomenon in which energy is transferred in the form of
electromagnetic waves between two separated objects with or without the existence of a medium
in between.
q F(T 4 Ta4 )
: Stefan-Boltzmann constant
F: Radiation view factor
: Emissivity
: Absorptivity
T, Ta: temperature of the object and ambient
Figure 13
In the automotive lamp application. Temperature distribution at the lens surface was measured 2
hours later after the bulb is turned on. Cavity radiation between bulb, reflector and lens is
considered.
NFX Application
Radiation heat transfer analysis of a automotive lamp
Figure 14
Linear and nonlinear heat transfer
Heat transfer analysis can also be classified as linear and nonlinear considering material
properties, heat transfer mechanisms and analysis conditions.
For linear heat transfer analysis, material properties are assumed to be constant and do not vary
according to temperature. However, most materials, especially metallic materials, have
properties (conductivity, specific heat, and density) that are temperature-dependent. Therefore,
when modeling and simulating temperature distribution for such materials, nonlinearities have to
be accounted.
For conduction and convection heat transfer, heat flux has linear relationship with temperature
difference. (see Figure 8, 10) However radiation heat transfer is a high order nonlinear
phenomenon due to T4 and Ta4 terms in the governing equation. (see Figure 13) Therefore,
nonlinear analysis is needed in radiation problems.
In addition, linear analysis consider loads and boundaries to be constant ( constant temperature
of heat source and environment). This may involve some assumptions and hypotheses.
Figure 15
Steady state and transient heat transfer
Furthermore heat transfer analysis can be classified as steady state and transient analysis.
Steady state analysis deals with problems in which the object and its surroundings reach
constant temperatures. At this state heat flow velocity and temperature distribution are steady
and do not vary according to time. While transient analysis deals with problems in which
temperatures within an object and surroundings changes as functions of time.
In normal situation, an object always passes Temperature
from transient state to steady state, imagine a
hot pan taken out from the oven and set aside to
cool down. At first temperature continuously
goes down until reaching some point, then
temperature stays almost constant at this low
temperature.
Therefore steady state analysis is used to Transient Steady-state
determine the final state and usually the
maximum temperature generated in a product Time
during the design. While transient analysis is
Figure 16 Temperature change in CPU
used to investigate the process in detail.
Another common application of transient analysis is to study drastic temperature changes based
on time (ex. iron, break pad, LED lamp, power semiconductors, etc).
Similar as transient response analysis in structural analysis, transient heat transfer analysis also
needs to set proper time intervals. If the time interval is too large, it may not capture the correct
temperature change. Because the temperature changes drastically in the beginning, we can use
short time intervals in the beginning and large time intervals near the steady state.
Fast initial
Temperature
temperature change
Time
Figure 18
For a chip in electronic device, thermal stress
occurs when electrical energy dissipated as heat,
and the chip starts to get hot.
For this kind of problem, temperature distribution
needs to be determined and further used as heat
load to calculate deformation and stress. Temperature distribution
Until here, we are discussing thermal problems in structural Model for heat transfer in
analysis, in which behavior of structural component itself is being structural analysis
studied under heating. For these problems fluid temperature of the Fluid at Ambient Temperature
environment is considered to be constant. Common structural
thermal problems are: heat generation, convection, radiation convection
problems and thermal stress problems in which thermal
deformation and stress caused by heat load are to be determined.
Solid Part
When temperature distribution of the surrounding fluid needs to be
studied; or the impact of the structural component on the
Model for heat transfer in
environment needs to be studied heat flow coupled analysis is CFD analysis
necessary. One of the most common application is to natural /
forced cooling of the system. Fluid Volume
In NFX, both structural heat transfer analysis and CFD solid / fluid
coupled analysis can be performed.
Solid Part
Figure 21
How to Choose Software for Thermal Analysis
We have well understood that FEA based verification tools are beneficial in many ways such as
reducing prototyping cost, discovering problems in early stage of the design, gaining insights on
thermal phenomenon which cannot be easily measured in physical tests, etc.
However, choosing the right software is more crucial in order to solve the exact problem as well
as to secure your project schedule and budget.
To choose the right CAE software, especially to adopt the simulation driven design approach in
a well established product design team, 4 major factors are to be considered here:
1) Solver capabilities
2) Easiness for modeling : CAD compatibility and integrated modeling tools
3) Speed and convenience of graphical output and reporting
4) Easiness to be learned and adopted in existing product design process
Solver capabilities
Considering the typical thermal problems introduced above, competent software should be able to
analyze
Conduction, convection and radiation problems
Temperature-dependent material properties and boundary conditions ( nonlinear heat transfer)
Time-dependent thermal effects ( transient thermal analysis)
Thermal-structural coupled analysis to calculate thermal stress and deformation
Thermal-fluid coupled analysis to calculate natural / forced convection problems
midas NFX solver has the capability to solve all the problems above and provide reliable results.
Some analysis features are dedicated to thermal analysis. Thermal sensor for example can be
put on any point or surface. It automatically detects when the system achieves steady state and
tells the solver to stop further calculation. This feature can reduce significantly calculation time of
transient thermal analysis.
Further more, as an all-in-one solution software, midas NFX is not only specific to thermal analysis,
but able to apply to structural, fluid dynamic and topology optimization analyses. This wide range of
analysis capabilities can be extremely beneficial to your product design team to check structural
and fluid flow problems in the products without having to work on multiple interfaces.
Especially, powered by multi-core solver and GPU calculation, midas NFX solver is featured with
its impressive analysis speed.
Easiness for modeling
New product are designed in CAD tools, many CAD tools are available on the market, and even in
one team different CAD tools are usually being used. Therefore, CAD compatibility is another
critical capacity to consider.
midas NFX is fully compatible with CAD, 13 most commonly used CAD formats are supported,
including Parasolid, CATIA V4/V5, UG, Pro/E, SolidWorks, Soild Edge, Inventor, STEP, IGES,
ACIS. Nastran format (.nas; .bdf) can also be import for analysis.
Models directly taken from CAD designs are usually with excessive detailing. Especially electronic
components, machine parts include many small holes, fillets, lines, faces which are not only
unnecessary but also will lead to inefficient analysis. Common practice is to clean up the geometry
model first before meshing and analyzing it.
midas NFX is equipped with automatic and manual simplification tools to handle effectively the
clean up work. With several clicks such detailing can be automatically detected, selected and
removed. This will save you considerably amount of time.
Furthermore, midas NFX is equipped with geometry creation and editing tools. Simple
modifications to the model can be directly done in midas NFX without going back to CAD tools.